RAPID RESPONSE OF HABITAT STRUCTURE AND ABOVE-GROUND CARBON STORAGE TO ALTERED FIRE REGIMES IN TROPICAL SAVANNA - MPG.PURE
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
Biogeosciences, 16, 1493–1503, 2019
https://doi.org/10.5194/bg-16-1493-2019
© Author(s) 2019. This work is distributed under
the Creative Commons Attribution 4.0 License.
Rapid response of habitat structure and above-ground carbon
storage to altered fire regimes in tropical savanna
Shaun R. Levick1,2,3 , Anna E. Richards2 , Garry D. Cook2 , Jon Schatz2 , Marcus Guderle1 , Richard J. Williams2 ,
Parash Subedi3 , Susan E. Trumbore1 , and Alan N. Andersen3
1 Max Planck Institute for Biogeochemistry, Hans-Knoell-Str. 10, 07745 Jena, Germany
2 CSIRO Land and Water, PMB 44, Winnellie, Darwin 0822 NT, Australia
3 Research Institute for the Environment and Livelihoods, Charles Darwin University, Casuarina NT 0909, Australia
Correspondence: Shaun R. Levick (shaun.levick@csiro.au)
Received: 13 April 2018 – Discussion started: 27 April 2018
Revised: 8 February 2019 – Accepted: 5 March 2019 – Published: 10 April 2019
Abstract. Fire regimes across the globe have been altered plications for habitat conservation, carbon sequestration, and
through changes in land use, land management, and climate emission reduction initiatives in the region.
conditions. Understanding how these modified fire regimes
impact vegetation structure and dynamics is essential for in-
formed biodiversity conservation and carbon management in
savanna ecosystems. We used a fire experiment at the Terri- 1 Introduction
tory Wildlife Park (TWP), northern Australia, to investigate
the consequences of altered fire regimes for vertical habi- Fire is an integral component of the functioning of savanna
tat structure and above-ground carbon storage. We mapped ecosystems, exerting top-down control on woody vegetation
vegetation three-dimensional (3-D) structure in high spatial structure (Bond and Keeley, 2005; Sankaran et al., 2005).
resolution with airborne lidar across 18 replicated 1 ha plots Savanna fires restrict vegetation vertical growth through a
of varying fire frequency and season treatments. We used “fire-trap” mechanism, whereby young trees are constrained
lidar-derived canopy height and cover metrics to extrapo- to low woody re-sprouts under high fire frequencies (Hig-
late field-based measures of woody biomass to the full extent gins et al., 2000; Freeman et al., 2017). A lengthening of the
of the experimental site (R 2 = 0.82, RMSE = 7.35 t C ha−1 ) fire-free interval allows trapped woody plants to grow above
and analysed differences in above-ground carbon storage and flame height, enabling them to reach mid- and upper-canopy
canopy structure among treatments. Woody canopy cover heights, with long-term consequences for size-class distri-
and biomass were highest in the absence of fire (76 % and bution and structural heterogeneity (Helm and Witkowski,
39.8 t C ha−1 ) and lowest in plots burnt late in the dry season 2012; Levick et al., 2015a).
on a biennial basis (42 % and 18.2 t C ha−1 ). Woody canopy Three-dimensional (3-D) heterogeneity of vegetation has
vertical profiles differed among all six fire treatments, with long been valued as a key factor promoting faunal di-
the greatest divergence in height classes < 5 m. The magni- versity through increased niche diversity and availability
tude of fire effects on vegetation structure varied along the (MacArthur and MacArthur, 1961; MacArthur, 1964). The
environmental gradient underpinning the experiment, with structural modifications that fires impart on savanna vegeta-
less reduction in biomass in plots with deeper soils. Our re- tion have been shown to impact both vertebrate (Woinarski
sults highlight the large extent to which fire management can et al., 2004) and invertebrate (Andersen et al., 2012) taxa.
shape woody structural patterns in savanna landscapes, even Fire-driven structural changes to savanna vegetation also
over time frames as short as a decade. The structural pro- have important implications for climate regulation, as sa-
file changes shown here, and the quantification of carbon re- vanna fires contribute significantly to atmospheric emissions
duction under late dry season burning, have important im- of greenhouse gases through biomass combustion (Hurst
et al., 1994; van der Werf et al., 2010). Despite the impor-
Published by Copernicus Publications on behalf of the European Geosciences Union.
1494 S. R. Levick et al.: Carbon consequences of altered fire regime tance of quantifying fire-induced changes to 3-D structure in Airborne lidar has a proven record in providing detailed 3-D savanna vegetation, current understanding of magnitudes and representations of savanna vegetation structure across time spatial patterns remains limited, and savanna fires represent and space (Smit et al., 2010; Levick et al., 2012, 2015b; large uncertainty in global vegetation models (Higgins et al., Goldbergs et al., 2018) but has yet to be used for assessing 2007; Scheiter et al., 2013). Gaining better understanding of vegetation biomass and structural diversity responses to ex- how different fire regimes impact savanna vegetation struc- perimental fires in savannas. ture is becoming increasingly urgent in the face of changing Northern Australia has a long history of savanna fire exper- climate and land-management conditions that are triggering iments (Williams et al., 2003), including the ongoing “Burn- variations in the timing, frequency, intensity, and duration of ing for Biodiversity” experiment at the Territory Wildlife fires in the tropical biome (Alencar et al., 2015). Park that has applied six fire treatments in three replicated Fire frequency in Australian savannas is particularly high, blocks since 2004 (Scott et al., 2010). Here we integrate with many regions burning twice in every 3 years on average field-based measurements of vegetation structure with air- (Beringer et al., 2014). Many of these fires occur late in the borne lidar to determine how variation in fire frequency and dry season, producing high-intensity burns that result in sim- season affects the 3-D habitat structure and above-ground plified vegetation structure (Bowman et al., 1988; Lehmann carbon storage of woody vegetation. Our specific aims are to et al., 2009; Ondei et al., 2017). There are widespread con- (i) explore how vegetation carbon storage and structural di- cerns that such fire regimes are linked to dramatic declines versity respond to increasing fire frequency and (ii) quantify in faunal populations, through the removal of ground layer the structural impact of late-season fires compared to early vegetation (Lawes et al., 2015; Legge et al., 2015; Woinarski season fires. We use airborne lidar data to provide greater et al., 2015). Methane and nitrous oxide emissions from sa- spatial coverage than can be achieved with field sampling vanna fires are included in Australia’s national greenhouse alone and to gain better understanding of how reliably lidar gas accounts and are responsible for approximately 3 % of could be used to assess savanna carbon dynamics in instances total accountable greenhouse gas emissions (Meyer et al., where field data may not be available or attainable. 2012). There is considerable interest in reducing the fre- quency and intensity of fires in northern Australia through strategic early dry season (April to July) burning in order 2 Methods to reduce both greenhouse gas emissions and certain com- ponents of biodiversity decline (Russell-Smith et al., 2013). 2.1 Study site and experimental design As such, the Australian government has implemented legisla- tion enabling landowners to claim carbon credits for reducing The Territory Wildlife Park is located 40 km south of Dar- greenhouse gas emissions from savanna fires through early win in Australia’s Northern Territory (Fig. 1). The vegeta- dry season burning (Carbon Farming Initiative – Emissions tion at the site is a mixed open forest and woodland savanna Abatement through Savanna Fire Management Methodology dominated by Eucalyptus miniata A.Cunn. ex Shauer, Euca- Determination 2015, Department of Environment and En- lyptus tetrodonta F.Muell., and Corymbia bleeseri (Blakely) ergy). Such changes to fire regimes in northern Australia are K.D.Hill and L.A.S.Johnson, with a grassy understorey also likely to increase carbon sequestration in the landscape dominated by Pseudopogonatherum contortum (Brongn.) (Murphy et al., 2010; Richards et al., 2012), although there A.Camus, Sarga intrans F.Muell. ex Benth., and Eriachne is currently no approved methodology for incorporating this triseta Nees ex Steud (Scott et al., 2010). The soils are rel- into the national accounts. While much attention is currently atively shallow (0.5 to 1 m deep) gravelly red earths (petro- being given to reducing the extent and frequency of late sea- ferric red Kandosol) (Isbell, 2002) of the Kay land system son fires in northern Australia, it is important to recognize within the Koolpinyah land surface group and have devel- that savannas have evolved with fire (Bond and Keeley, 2005; oped predominantly from deeply weathered sandstones, silt- Durigan and Ratter, 2016) and excluding fire would be detri- stones, and shales (Wood et al., 1985). The climate is wet- mental to certain savanna specialists that favour more open dry tropical with greater than 90 % of annual rainfall (mean and grassy habitat. The challenge is finding the best mix of 1401 mm) falling in the wet season from November to April, patches of different regimes across connected landscapes. and mean monthly maximum and minimum temperatures be- Understanding of how different fire regimes impact habi- tween 33.1 and 20.9 ◦ C (Bureau of Meteorology, Common- tat structure and carbon dynamics in tropical savannas can wealth of Australia). be enhanced through detailed 3-D measurements of vegeta- The fire experiment consists of 18 1 ha plots grouped into tion structure at sites subject to long-term, replicated experi- three blocks (A, B, C) arranged along a north–south tran- mental fire treatments. Traditional field-based inventory tech- sect (Fig. 1). Soils are deeper at the southern end, and the C niques are limited in their ability to quantify 3-D structure, block has higher soil moisture given its proximity to a small but light detection and ranging (lidar) can now achieve this drainage line. Six fire treatments were randomly assigned to with high accuracy and precision in a repeatable and transfer- each block at the start of the experiment: unburnt plots (U); able manner (Lefsky et al., 2002; Levick and Rogers, 2008). plots burnt at fire return intervals of 1 (E1), 2 (E2), 3 (E3), Biogeosciences, 16, 1493–1503, 2019 www.biogeosciences.net/16/1493/2019/
S. R. Levick et al.: Carbon consequences of altered fire regime 1495
Figure 1. Location and experimental design of the Territory Wildlife Park fire manipulation experiment. Treatments were fist implemented
in 2004. Soil depth and soil moisture increase from the northern to southern blocks.
and 5 (E5) years in the early dry season (E2); and plots burnt in height were recorded. The location of each individual
every 2 years (L2) in the late dry season (Table 1). Prior plant was recorded to 0.3 m accuracy using a differential
to implementation of the burning treatments in 2004, all ar- GPS with post-processing (Trimble Inc.). Tree heights were
eas had been unburnt for at least 14 years when fire records recorded with a standard height pole (plants < 8 m) or cli-
started (except for a fire in 1992 and again in 2000 in the A nometer (plants > 8 m), and stem diameter was recorded at
block only). 1.3 m with a diameter tape for all woody species except for
During each experimental burn, fire intensity was esti- the multi-stemmed shrubs Calytrix exstipulata and Exocar-
mated using the established relationship between rate of pos latifolius, in which case diameter was recorded at the
spread and fuel load (Williams et al., 1998). Rate of fire stem base (0.1 m above the ground). Above-ground biomass
spread was determined from thermocouples positioned 5 cm was calculated for each individual tree using the equation de-
above the soil surface in the flaming combustion zone, linked veloped by Williams et al. (2005):
to buried electronic stop watches. Six timers were used in
each 1 ha plot, arranged in a series of equilateral triangles, ln ABG = −2.0596 + 2.1561(ln D) + 0.1362(ln H )2 , (1)
with 10 m sides. The rate of fire spread was also determined
by observers using stop watches, manually recording the whereby AGB is above-ground biomass (kg), D is stem
time of arrival at the points where the electronic watches diameter (m), and H is tree height (m). Individual tree
were positioned. All points were marked by star pickets biomasses were then summed for each 30 m × 30 m subplot.
and flagging tape. Fuel loads were determined prior to each Estimated biomass values were converted to carbon terms on
fire by direct harvest and weighing. A total of 10 replicate a per hectare basis assuming 50 % of biomass was carbon
0.5 m × 0.5 m fuel samples were cut for each plot. Fuel heat (t C ha−1 ). This approach did not consider the contribution
content was assumed to be 20 000 kJ kg−1 dry weight. of small (< 2 m) multi-stemmed shrubs to the carbon pool.
2.3 Airborne lidar surveying and processing
2.2 Field-based estimation of above-ground woody
biomass We mapped 150 ha of the study area with airborne lidar in
June 2013, 9 years after the start of the experiment. The air-
In each of the 18 plots, two 30 m × 30 m subplots were es- borne survey was conducted by Airborne Research Australia
tablished at the north-west and south-east corners, at least (ARA) with a full-waveform lidar sensor (RIEGL LMS-
10 m away from plot edges. In each subplot the species iden- Q560) operated from a light fixed-wing aircraft (Diamond
tity, location, height, and diameter of all woody plants >2 m Aircraft ECO Dimona). Flight lines with > 50 % overlap
www.biogeosciences.net/16/1493/2019/ Biogeosciences, 16, 1493–1503, 2019
1496 S. R. Levick et al.: Carbon consequences of altered fire regime
Table 1. Fire regime characteristics of the Territory Wildlife Park experimental site. Data are for the period 2004–2013. Fire intensity values
are the mean and standard error over the course of the experiment.
Treatment Season Frequency (year) Intensity (kW m−1 ) Times burnt
E1 June 1 589 ± 144 9
E2 June 2 929 ± 20 4
E3 June 3 424 ± 26 3
E5 June 5 295 ± 69 2
L2 October 2 1644 ± 131 5
U n/a 0 0 0
n/a – not applicable
were used to achieve double coverage of the plots (average the most robust model (in terms of explanatory power and
flying height 300 m a.g.l., swath width 250 m, line spacing RMSE) across the full extent of the airborne lidar coverage to
125 m), and the RIEGL LMS-Q560 was operated at 240 kHz examine the effects of fire treatment on above-ground woody
and 135 lines per second. Slow flying speed of less than biomass.
40 m s−1 ensured high point densities along track, with an av-
erage return density of 22.28 m2 and an average pulse spac- 2.5 Assessment of treatment effects
ing of 0.21 m.
Raw lidar data were processed with RiANALYZE (RIEGL We digitally distributed six 30 m × 30 m subplots in each
Laser Measurement Systems GmbH) for decomposing the fire plot for statistical comparison of treatment effects. We
full waveforms into discrete returns. The ARA RASP open- used a linear mixed-effect modelling approach, with Gaus-
source software (RASP version 0.98: manual, code, and ex- sian residual variance, to test the significance of fire treat-
ecutables available from ARA on request) was used to ori- ment on woody canopy cover, canopy height, and above-
entate the point cloud to Cartesian coordinates and output ground biomass. The models were implemented in R (R Core
the geolocated point cloud in the American Society for Pho- Team, 2018) with the nlme package (Pinheiro et al., 2018).
togrammetry and Remote Sensing (ASPRS) standard LAS Maximum likelihood (ML) was used to fit the models, with
format. All further point-cloud processing tasks were con- subplots included as a random effect nested within fire treat-
ducted with the LAStools suite of processing scripts (Rapid- ments. Fixed effects were fire treatment (E1, E2, E3, E4, E5,
lasso GmbH). The last returns were classified into ground L2, U), block position (A, B, C), and the interaction between
and non-ground points for bare-earth extraction. A digital ter- fire treatment and block position. Models were generated for
rain model (DTM) was constructed from ground returns us- all possible combinations of fixed effects, together with a null
ing a triangulated irregular network approach (TIN) at 0.25 m model consisting of only the random effects of the quadrat lo-
resolution. The DTM was used to normalize the z coordinate cations. Akaike information criterion (AIC) scores for each
of vegetation returns to height above ground level (Fig. 2). of the models were compared to identify the most parsimo-
nious model.
2.4 Upscaling above-ground woody biomass estimates The impact of fire treatment on vegetation vertical profile
with airborne lidar distribution was explored by plotting the mean and 95 % con-
fidence interval of lidar returns per 0.5 m height class. Statis-
The normalized airborne lidar returns were clipped to the tical significance of treatment effects was tested pairwise on
spatial extent of each field-measured 30 m × 30 m subplot. a per height class basis using a paired t test.
Using the lascanopy tool within LAStools, we extracted a
suite of 14 ecologically meaningful metrics describing veg-
etation structure from the point cloud: mean canopy height 3 Results
(MCH), quadratic mean canopy height (QMCH), canopy
3.1 Estimation of above-ground woody biomass from
cover > 1 m (COV1), canopy cover > 10 m (COV10), canopy
airborne lidar
density (DENS), kurtosis (KUR), skewness (SKE), standard
deviation (SDE), canopy relief ratio (CRR), and a series of Airborne lidar proved valuable for upscaling woody biomass
height quantiles (Q10 , Q25 , Q50 , Q75 , Q90 ). Using these 14 measurements from the field plots to the full extent of the
metrics as explanatory variables, we ran step-wise multiple fire experiment (Fig. 3). Only three woody canopy struc-
linear regression with Akaike information criterion (AIC) tural variables were retained in the step-wise linear regres-
minimization against the field-estimated biomass to iden- sion procedure: mean canopy height (MCH), total canopy
tify the variables with the most explanatory power and used cover (Cov1m), and overstorey canopy cover (Cov10m):
them to construct a lidar-based biomass model. We applied
Biogeosciences, 16, 1493–1503, 2019 www.biogeosciences.net/16/1493/2019/
S. R. Levick et al.: Carbon consequences of altered fire regime 1497
Figure 2. Cross section through the normalized high-resolution lidar point cloud (a) and aerial view of rasterized canopy height model
(CHM) interpolation (b). The lidar point cloud provided excellent representation of both the vertical and horizontal structures of vegetation
across the site.
AGB = −6.524 + (−0.794Cov10m) + (−0.345Cov1m)
+ (14.881MCH). (2)
The distribution of model residuals showed no spatial
trend or relationship with the fire treatment. The degree of
residual error (RMSE = 7.35 t C ha−1 ) provided acceptable
confidence for inclusion of modelled biomass values in fur-
ther analyses.
3.2 Effects of fire regime on woody canopy cover and
above-ground biomass
Canopy cover decreased along the experimental gradient
of fire frequency and season, ranging from about 75 %
(SE = 1.7) in unburnt plots to 45 % (SE = 2.3) in late season
biennial plots (Fig. 4a). These differences in canopy cover
translated into similar patterns of biomass variation across
the experiment (Fig. 4b). The highest within-treatment vari- Figure 3. Relationship between field-estimated above-ground
ability for both cover and biomass was found in the early biomass and estimates predicted from airborne lidar metrics.
season annual plots (E1). Open green circles represent individual subplots (30 m × 30 m); the
The best model explaining variation in both woody cover dashed line shows the linear fit.
and biomass was one in which fire treatment, block posi-
tion, and the interaction between them was included (Ta-
ble 2). Model performance was poorer when the interac- (1AIC = 73.34 vs. 51.70). These results point to an impor-
tion term was excluded (1AIC = 48.91, 32.98, and 39.29 for tant source of environmental variation arising from block po-
woody cover, height, and biomass respectively). When ex- sition, which represents a gradient in soil depth and moisture
planatory variables were considered independently, fire treat- availability across the experimental site.
ment was more influential than block position on variation When we consider the spectrum of increasing fire intensity
in woody cover (1AIC = 70.25 vs. 90.67), but not for mean occurring across the experimental treatments, we found that
canopy height (1AIC = 59.96 vs. 48.93) or woody biomass correlations between the reductions in above-ground biomass
www.biogeosciences.net/16/1493/2019/ Biogeosciences, 16, 1493–1503, 20191498 S. R. Levick et al.: Carbon consequences of altered fire regime
Table 2. Linear mixed model results of lidar-estimated canopy cover, mean height, and above-ground woody biomass.
Model terms AIC cover 1AIC cover AIC height 1AIC height AIC biomass 1AIC biomass
Fire treatment · block 806.14 0.00 406.47 0.00 864.51 0.00
Fire treatment + block 855.05 48.91 439.45 32.98 903.81 39.29
Fire treatment 876.59 70.45 466.43 59.96 937.87 73.34
Block 896.81 90.67 455.40 48.93 916.23 51.70
Null model 915.34 109.20 477.03 70.56 944.09 79.57
and fire intensity decreased along the soil depth and mois- the Territory Wildlife Park fire experiment. Fire effects were
ture availability gradient (Fig. 5). In carbon terms, the early most pronounced at the extremes of the experimental spec-
biennial fires on average caused a reduction of 10 t C ha−1 trum, with the highest cover and biomass occurring under
compared to unburnt plots, whereas late biennial fires almost complete fire exclusion and the lowest values of woody
doubled that reduction to 19 t C ha−1 . canopy structure obtained under biennial late season burn-
ing. The directionality of these trends was persistent across
3.3 Fire effects on vertical habitat structure the underlying gradient of increasing soil depth and mois-
ture, but the magnitude and slope of the effects was greater
In addition to the observed patterns in woody canopy cover in the A and B blocks with shallower, drier soils (Fig. 5). The
and above-ground biomass, our lidar-based assessment also lower magnitude of carbon reduction in the lower-lying C
revealed substantial variation in canopy height profile distri- block likely stems from the sparse herbaceous cover in these
butions, derived from the number of lidar returns from differ- plots, which results in patchy, low-intensity fires.
ent height levels (Fig. 6). Most profiles were bimodal, with Recent research into woody biomass trends in the re-
a peak at 1–2 m in height and a smaller peak at 10–15 m. gion (from long-term field monitoring plots) indicate that
The clearest bimodal response was found in the early season woody biomass has been relatively stable over decadal pe-
triennial burns (Fig. 6c), whereas early season annual and 5- riods, with minor evidence of woody thickening, and that
year burn profiles were more uniform (Fig. 6a, d). biomass is negatively correlated with fire frequency (Mur-
Keeping fire frequency constant (biennial) and exploring phy et al., 2013). However, as Fensham et al. (2017) note,
the effects of fire season highlighted the large influence of a key finding emerging from that regional study was that
late season versus early season burns (Fig. 7a). Compared the observed decreases in tree biomass following severe fires
with no fire, early season biennial fires significantly reduced were not driven by mortality of individual trees, but rather
canopy below 5.5 m and late season biennial fire reduced by decreases in the rates of biomass accumulation of surviv-
canopy even further up to the 9.5 m height class (p < 0.05, ing trees. We do not have repeated individual tree data in our
Fig. 7b), generating a vertical profile similar in shape but study to directly corroborate this finding, but the patterns of
with lower frequency of occurrence. The late season fire pro- reduced cover throughout the height profile do suggest mor-
file contained significantly less canopy in nearly all height tality and the consumption of trees by fire, rather than just
classes compared to the unburnt (p < 0.05, Fig. 7a, b), but reduction in growth rates.
the most marked effects were in the lower height classes, Similar investigations in southern African savannas have
which represent the shrub layer and the recruitment zone. found that fire frequency itself had little bearing on woody
cover, but that the presence of fire alone was a stronger pre-
4 Discussion dictor of reduced woody cover (Devine et al., 2015). In our
study, however, we found that cover and biomass were re-
Airborne lidar provided direct measures of canopy cover and duced as fire frequency increased (Fig. 4), with the excep-
height distribution, and the derived metrics successfully pre- tion to the trend being the early biennial fires (E2), which
dicted field-based estimates of above-ground biomass. The had a slightly larger impact on structure than the early an-
synoptic view that airborne lidar provided enabled us to map nual (E1) fires. The experimental design incorporates fire fre-
changes in biomass under different fire regimes, in addition quency and season, but the net result of these components
to exploring differences in vegetation vertical profiles across of the fire regime is fire intensity, which is the stronger de-
the full expanse of the fire experiment. termining factor of vegetation structural change (Williams
et al., 1999; Furley et al., 2008). Of all the fire treatments,
4.1 Carbon storage consequences of altered fire the biennial burns had the highest mean intensities of 929
regimes and 1664 kW m−1 for early season and late season fire re-
spectively. These intensities are still low compared to those
A total of 10 years of experimental burning imparted large of large late season fires in northern Australia and reflect
structural differences in woody canopy across the plots of
Biogeosciences, 16, 1493–1503, 2019 www.biogeosciences.net/16/1493/2019/S. R. Levick et al.: Carbon consequences of altered fire regime 1499
Figure 5. Density plot showing the relationship between increasing
fire intensity reduction in woody carbon storage, relative to unburnt
plots, for the A (blue), B (red), and C (green) blocks.
There is increasing interest in understanding the effect of
different fire regimes on carbon stored in Australian savan-
nas (Murphy et al., 2013; Cook et al., 2015) and recent stud-
ies have shown higher carbon stocks in dead organic matter
under lower fire frequencies (Cook et al., 2016). At the Terri-
tory Wildlife Park fire plots the early biennial fire caused a re-
duction of 10 t C ha−1 on average compared to unburnt plots,
whereas late biennial fires almost doubled that average re-
duction to 19 t C ha−1 (Fig. 5). These patterns are consistent
with the trend of lower greenhouse gas emissions under early
dry season fires, relative to late fires (Meyer et al., 2012), and
point to the importance of available fuel load and its char-
acteristics (greater herbaceous volume and lower moisture
content late in the dry season) in understanding fire-induced
structural change in savannas. This is further emphasized by
the variation in response to fire along the environmental gra-
dient of the experimental site.
Murphy et al. (2013) suggested that the moderation of fire
regimes in northern Australia is likely to increase carbon
storage in woody biomass, but the extent to which woody
biomass can increase in these savannas is highly uncertain.
Figure 4. Relationship between fire treatments and (a) woody
Our results reduce some of this uncertainty, by providing
canopy cover and (b) woody biomass. Fire treatments ordered ac-
cording to increasing fire intensity. Green dots in (b) indicate field-
quantification of the degree to which carbon stored in unburnt
measured values derived from 30 m × 30 m subplots. plots deviates from a range of different fire frequencies.
4.2 Shifts in vegetation vertical profile distribution
under altered fire regimes
the small scale of the experimental plots. Nonetheless, de- Different fire regimes imparted a diverse array of vertical
spite lower intensities across the board compared to larger structural profiles on woody vegetation. Although woody
experiments like those obtained at the Kapalga experiment, canopy cover and above-ground biomass displayed subtle re-
our findings are in agreement with the diminished basal ar- sponses among the early season fire frequency treatments,
eas observed there under very-high-intensity late season fires we found that each fire regime generated a relatively unique
(Andersen et al., 2003). niche space in terms of vertical profile distribution. These
www.biogeosciences.net/16/1493/2019/ Biogeosciences, 16, 1493–1503, 20191500 S. R. Levick et al.: Carbon consequences of altered fire regime Figure 6. Effects of fire regime on vertical habitat structure determined from the frequency of airborne lidar returns. Solid black lines are the mean frequency distribution of lidar returns, and the green bands indicate the 95 % confidence interval. Figure 7. Effect of biennial fire season on woody vertical profile structure with unburnt treatment in green, early season biennial in blue, and late season biennial in red. (a) Dots and error bars represent the mean and 95 % confidence interval for the 30 m × 30 m subplots (n = 18). Some returns from lower height classes may have arisen from herbaceous material. (b) Pairwise comparison of differences in means on a per height class basis. U–L2 is unburnt vs. late season biennial, U–E2 is unburnt vs. early season biennial, and L2–E2 is late season vs. early season biennial. p values are shaded with values < 0.05 in red. Biogeosciences, 16, 1493–1503, 2019 www.biogeosciences.net/16/1493/2019/
S. R. Levick et al.: Carbon consequences of altered fire regime 1501
niches were most divergent in the understorey height classes proach should ideally be complimented with time-series
(< 5 m). Tracking these profiles over time into the future analyses of before and after fire events to better constrain the
might reveal increased height of divergence as cohorts grow mechanisms underpinning structural change.
taller. Alternatively, these understorey height curves may
represent stable persistent equilibrium re-sprout heights that
define the optimum for re-sprouts that are able to persist
5 Conclusions
within the flame zone under a particular fire regime (Free-
man et al., 2017). We quantified the magnitude of above-ground carbon reduc-
These vertical profile findings highlight the powerful role tion under different regimes by integrating airborne lidar,
that fire management can play in shaping three-dimensional field surveys, and an ongoing fire regime experiment. Our re-
habitat structure in ecosystems. The challenge this presents sults highlight the impact of late season burning on both car-
to land managers is deciding which of this range of profiles bon storage and canopy vertical profile structure. Clear rela-
is optimal for their specific management objectives. We still tionships between biodiversity and fire regimes have proven
lack explicit understanding of how different organisms utilize difficult to establish in savannas, despite many attempts at
three-dimensional space, and it is increasingly evident that linking floral and faunal diversity directly to fire regime pat-
no one profile is optimal. Mid-storey shrubs and trees provide terns. The range of vertical profile responses that we have
key food resources for birds and small mammals, and high illustrated here under different experimental fire treatments
ground cover reduces predation risk by feral cats (Davies could hold the key to unlocking stronger links between fire
et al., 2016). Conversely, habitat simplification through late management and biodiversity responses. High-resolution li-
season burning was found to promote longer-term abundance dar can expose the structural consequences of different man-
of frilled-neck lizards in Kakadu National Park, despite high agement actions and make them more easily accessible for
initial direct mortality rates (Corbett et al., 2003; Andersen integration with biodiversity and ecosystem process studies.
et al., 2005). As such, it is likely that a mix of patches at
the landscape scale, spanning a diverse range of vertical pro-
files, is needed from a wildlife conservation perspective. The Data availability. Data are accessible via the CSIRO Data Portal
relative proportions and spatial arrangement of these patches (https://data.csiro.au/dap, last access: 2 April 2019).
needs targeted and deeper investigation.
4.3 Limitations and future directions Author contributions. SRL, AER, GDC, SET, and ANA conceived
the study. JS, PS, and AER conducted the field research. RJW and
Our findings in this study provide quantification of the mag- JS collected the fire intensity data. SRL and MG processed the li-
nitudes of fire regime effects on woody structure in a tropical dar data and conducted the analyses. All authors contributed to the
savanna. When generalizing to other savanna regions how- writing of the paper.
ever, the following limitations should be to be taken into con-
sideration. First, prior to the establishment of the TWP fire
experiment in 2004 the vegetation was unburnt since 1990. Competing interests. The authors declare that they have no conflict
In these tropical landscapes 14 years of fire exclusion is rare, of interest.
so the starting conditions are atypical.
Second, despite the good results obtained in upscaling
Acknowledgements. This project was jointly funded through
field-based woody biomass estimates with airborne lidar
CSIRO, Charles Darwin University, and the Australian Govern-
(Fig. 3), future efforts should focus on reducing the level of ment’s National Environmental Science Programme (NESP). We
uncertainty in the lidar biomass model. Greater confidence in acknowledge the Northern Territory Government and the Territory
biomass and carbon prediction could be achieved by turning Wildlife Park for establishing and maintaining the long-term fire
to individual tree-based segmentation approaches. Develop- experimental site. Jorg Hacker and Airborne Research Australia
ments in terrestrial lidar in particular show great promise for (ARA) are thanked for their airborne surveying of the site.
providing individual tree volumes and biomass estimates that Shaun R. Levick was supported by grant number 01DR14010 from
can be scaled, together with their uncertainties, to plot and the BMBF (FIREBIODIV project).
landscape scales (Calders et al., 2014; Levick et al., 2016;
Singh et al., 2018). Furthermore, the rich 3-D models that The article processing charges for this open-access
terrestrial lidar provide will open up new avenues for explor- publication were covered by the Max Planck Society.
ing actual 3-D structural metrics.
Last, our analyses in this study rely on differences between
Review statement. This paper was edited by Christopher Still and
treatments at a single point in time to infer the mechanisms
reviewed by three anonymous referees.
underpinning woody structural modification. Although typ-
ical for this type of investigation, the single-time-point ap-
www.biogeosciences.net/16/1493/2019/ Biogeosciences, 16, 1493–1503, 20191502 S. R. Levick et al.: Carbon consequences of altered fire regime
References Fensham, R. J., Freeman, M. E., Laffineur, B., Macdermott, H.,
Prior, L. D., and Werner, P. A.: Variable rainfall has a greater
effect than fire on the demography of the dominant tree in a semi-
Alencar, A. A., Brando, P. M., Asner, G. P., and Putz, F. E.: Land- arid Eucalyptus savanna, Austral Ecol., 42, 772–782, 2017.
scape fragmentation, severe drought, and the new Amazon forest Freeman, M. E., Vesk, P. A., Murphy, B. P., Cook, G. D., Richards,
fire regime, Ecol. Appl., 25, 1493–1505, 2015. A. E., and Williams, R. J.: Defining the fire trap: Extension of the
Andersen, A. N., Cook, G. D., and Williams, R. J. (Eds.): Fire persistence equilibrium model in mesic savannas, Austral Ecol.,
in tropical savannas: the Kapalga experiment, Springer-Verlag, 42, 890–899, 2017.
New York, USA, https://doi.org/10.1007/b97225, 2003. Furley, P. A., Rees, R. M., Ryan, C. M., and Saiz, G.: Savanna burn-
Andersen, A. N., Cook, G. D., Corbett, L. K., Douglas, M. M., ing and the assessment of long-term fire experiments with partic-
Eager, R. W., Russel-Smith, J., Setterfield, S. A., Williams, ular reference to Zimbabwe, Prog. Phys. Geogr., 32, 611–634,
R. J., and Woinarski, J. C. Z.: Fire frequency and biodiver- 2008.
sity conservation in Australian tropical savannas: implications Goldbergs, G., Levick, S. R., Lawes, M., and Edwards, A.: Hi-
from the Kapalga fire experiment, Austral. Ecol., 30, 155–167, erarchical integration of individual tree and area-based ap-
https://doi.org/10.1111/j.1442-9993.2005.01441.x, 2005. proaches for savanna biomass uncertainty estimation from
Andersen, A. N., Woinarski, J. C. Z., and Parr, C. L.: Savanna burn- airborne LiDAR, Remote Sens. Environ., 205, 141–150,
ing for biodiversity: Fire management for faunal conservation in https://doi.org/10.1016/j.rse.2017.11.010, 2018.
Australian tropical savannas, Austral Ecol., 37, 658–667, 2012. Helm, C. V. and Witkowski, E. T. F.: Characterising wide spatial
Beringer, J., Hutley, L. B., Abramson, D., Arndt, S. K., Briggs, P., variation in population size structure of a keystone African sa-
Bristow, M., Canadell, J. G., Cernusak, L. A., Eamus, D., Ed- vanna tree, Forest Ecol. Manag., 263, 175–188, 2012.
wards, A. C., Evans, B. J., Fest, B., Goergen, K., Grover, S. P., Higgins, S. I., Bond, W. J., and Trollope, W. S. W.: Fire, resprouting
Hacker, J., Haverd, V., Kanniah, K., Livesley, S. J., Lynch, A., and variability: a recipe for grass-tree coexistence in savanna, J.
Maier, S., Moore, C., Raupach, M., Russell-Smith, J., Scheiter, Ecol., 88, 213–229, 2000.
S., Tapper, N. J., and Uotila, P.: Fire in Australian savannas: from Higgins, S. I., Bond, W. J., February, E. C., Bronn, A., Euston-
leaf to landscape, Glob. Change Biol., 21, 62–81, 2014. Brown, D. I. W., Enslin, B., Govender, N., Rademan, L.,
Bond, W. and Keeley, J.: Fire as a global “herbivore”: the ecology O’Regan, S., Potgieter, A. L. F., Scheiter, S., Sowry, R., Trollope,
and evolution of flammable ecosystems, Trends Ecol. Evol., 20, L., and Trollope, W. S. W.: Effects of four decades of fire manip-
387–394, 2005. ulation on woody vegetation structure in Savanna, Ecology, 88,
Bowman, D. M. J. S., Wilson, B. A., and Hooper, R. J.: Response 1119–1125, 2007.
of Eucalyptus Forest and Woodland to Four Fire Regimes at Hurst, D. F., Griffith, D. W. T., and Cook, G. D.: Trace gas emis-
Munmarlary, Northern Territory, Australia, J. Ecol., 76, 215–232, sions from biomass burning in tropical Australian savannas, J.
https://doi.org/10.2307/2260465, 1988. Geophys. Res.-Atmos., 99, 16441–16456, 1994.
Calders, K., Newnham, G., Burt, A., Murphy, S., Raumonen, P., Isbell, R. F.: The Australian Soil Classification, CSIRO Publishing,
Herold, M., Culvenor, D., Avitabile, V., Disney, M., Armston, J., Collingwood, 2002.
and Kaasalainen, M.: Nondestructive estimates of above-ground Lawes, M. J., Murphy, B. P., Fisher, A., Woinarski, J. C. Z., Ed-
biomass using terrestrial laser scanning, Methods Ecol. Evol., 6, wards, A. C., and Russell-Smith, J.: Small mammals decline with
198–208, 2014. increasing fire extent in northern Australia: evidence from long-
Cook, G. D., Liedloff, A. C., Cuff, N. J., Brocklehurst, P. S., and term monitoring in Kakadu National Park, Int. J. Wildland Fire,
Williams, R. J.: Stocks and dynamics of carbon in trees across 24, 712–722, 2015.
a rainfall gradient in a tropical savanna, Austral Ecol., 40, 845– Lefsky, M., Cohen, W., Parker, G., and Harding, D.: Lidar Remote
856, 2015. Sensing for Ecosystem Studies, BioScience, 52, 19–30, 2002.
Cook, G. D., Meyer, C. P. M., Muepu, M. L., and Liedloff, A. C.: Legge, S., Garnett, S., Maute, K., Heathcote, J., Murphy, S.,
Dead organic matter and the dynamics of carbon and greenhouse Woinarski, J. C. Z., and Astheimer, L.: A Landscape-Scale,
gas emissions in frequently burnt savannas, Int. J. Wildland Fire, Applied Fire Management Experiment Promotes Recovery of
25, 1252–1263, 2016. a Population of the Threatened Gouldian Finch, Erythrura
Corbett, L. K., Andersen, A. N., and Müller, W. J.: Ter- gouldiae, in Australia’s Tropical Savannas, PLoS ONE, 10,
restrial vertebrates, in: Fire in tropical savannas: the Ka- e0137997, https://doi.org/10.1371/journal.pone.0137997, 2015.
palga experiment, edited by: Andersen, A. N., Cook, G. D., Lehmann, C. E. R., Prior, L. D., and Bowman, D. M. J. S.: Fire
Williams, R. J., Springer-Verlag, New York, USA, 126–152, controls population structure in four dominant tree species in a
https://doi.org/10.1007/b97225, 2003. tropical savanna, Oecologia, 161, 505–515, 2009.
Davies, H. F., McCarthy, M. A., Firth, R. S. C., Woinarski, J. C. Z., Levick, S. R. and Rogers, K. H.: Structural biodiversity monitoring
Gillespie, G. R., Andersen, A. N., Geyle, H. M., Nicholson, E., in savanna ecosystems: Integrating LiDAR and high resolution
and Murphy, B. P.: Top-down control of species distributions: imagery through object-based image analysis, in: Object-based
feral cats driving the regional extinction of a threatened rodent in image analysis: spatial concepts for knowledge-driven remote
northern Australia, Divers. Distrib., 23, 272–283, 2016. sensing applications, edited by: Blaschke, T., Lang, S., and Hay,
Devine, A. P., Stott, I., McDonald, R. A., and Maclean, I. M. D.: G., 477–492, Springer-Verlag, Berlin, 2008.
Woody cover in wet and dry African savannas after six decades Levick, S. R., Asner, G. P., and Smit, I. P. J.: Spatial patterns in the
of experimental fires, J. Ecol., 103, 473–478, 2015. effects of fire on savanna vegetation three-dimensional structure,
Durigan, G. and Ratter, J. A.: The need for a consistent fire policy Ecol. Appl., 22, 2110–2121, 2012.
for Cerrado conservation, J. Appl. Ecol., 53, 11–15, 2016.
Biogeosciences, 16, 1493–1503, 2019 www.biogeosciences.net/16/1493/2019/S. R. Levick et al.: Carbon consequences of altered fire regime 1503 Levick, S. R., Baldeck, C. A., and Asner, G. P.: Demographic lega- Scheiter, S., Langan, L., and Higgins, S. I.: Next-generation dy- cies of fire history in an African savanna, Funct. Ecol., 29, 131– namic global vegetation models: learning from community ecol- 139, 2015a. ogy, New Phytol., 198, 957–969, 2013. Levick, S. R., Setterfield, S., Rossiter-Rachor, N., Hutley, L., Scott, K. A., Setterfield, S. A., Douglas, M. M., and Andersen, MacMaster, D., and Hacker, J.: Monitoring the Distribu- A. N.: Fire tolerance of perennial grass tussocks in a savanna tion and dynamics of an invasive grass in tropical sa- woodland, Austral Ecol., 35, 858–861, 2010. vanna using airborne LiDAR, Remote Sens., 7, 5117–5132, Singh, J., Levick, S. R., Guderle, M., Schmullius, C., and Trumbore, https://doi.org/10.3390/rs70505117, 2015b. S. E.: Variability in fire-induced change to vegetation physiog- Levick, S. R., Hessenmöller, D., and Schulze, E.-D.: Scaling wood nomy and biomass in semi-arid savanna, Ecosphere, 9, e02514, volume estimates from inventory plots to landscapes with air- https://doi.org/10.1002/ecs2.2514, 2018. borne LiDAR in temperate deciduous forest, Carbon Balance and Smit, I. P. J., Asner, G. P., Govender, N., Kennedy-Bowdoin, T., Management, 11, 7, https://doi.org/10.1186/s13021-016-0048-7, Knapp, D. E., and Jacobson, J.: Effects of fire on woody vege- 2016. tation structure in African savanna, Ecol. Appl., 20, 1865–1875, MacArthur, R. H.: Environmental factors affecting bird species di- 2010. versity, Am. Nat., 98, 387–397, 1964. van der Werf, G. R., Randerson, J. T., Giglio, L., Collatz, G. MacArthur, R. H. and MacArthur, J. W.: On bird species diversity, J., Mu, M., Kasibhatla, P. S., Morton, D. C., DeFries, R. S., Ecology, 42, 594–598, 1961. Jin, Y., and van Leeuwen, T. T.: Global fire emissions and the Meyer, C. P., Cook, G. D., Reisen, F., Smith, T. E. L., Tattaris, contribution of deforestation, savanna, forest, agricultural, and M., Russell-Smith, J., Maier, S. W., Yates, C. P., and Wooster, peat fires (1997–2009), Atmos. Chem. Phys., 10, 11707–11735, M. J.: Direct measurements of the seasonality of emission fac- https://doi.org/10.5194/acp-10-11707-2010, 2010. tors from savanna fires in northern Australia, J. Geophys. Res., Williams, R., Gill, A., and Moore, P.: Seasonal changes in fire be- 117, D20305, https://doi.org/10.1029/2012JD017671, 2012. haviour in a tropical savanna in northern Australia, Int. J. Wild- Murphy, B. P., Russell-Smith, J., and Prior, L. D.: Frequent fires re- land Fire, 8, 227–239, 1998. duce tree growth in northern Australian savannas: implications Williams, R. J., Cook, G. D., Gill, A. M., and Moore, P. H. R.: Fire for tree demography and carbon sequestration, Glob. Change regime, fire intensity and tree survival in a tropical savanna in Biol., 16, 331–343, 2010. northern Australia, Austral Ecol., 24, 50–59, 1999. Murphy, B. P., Lehmann, C. E. R., Russell-Smith, J., and Lawes, Williams, R. J., Woinarski, J. C. Z., and Andersen, A. N.: Fire ex- M. J.: Fire regimes and woody biomass dynamics in Australian periments in northern Australia: contributions to ecological un- savannas, J. Biogeogr., 41, 133–144, 2013. derstanding and biodiversity conservation in tropical savannas, Ondei, S., Prior, L. D., Williamson, G. J., Vigilante, T., and Bow- Int. J. Wildland Fire, 12, 391–402, 2003. man, D. M. J. S.: Water, land, fire, and forest: Multi-scale deter- Williams, R. J., Zerihun, A., Montagu, K. D., Hoffman, M., Hutley, minants of rainforests in the Australian monsoon tropics, Ecol. L. B., and Chen, X.: Allometry for estimating aboveground tree Evol., 7, 1592–1604, 2017. biomass in tropical and subtropical eucalypt woodlands: towards Pinheiro, J., Bates, D., DebRoy, S., Sarkar, D., and R Core Team: general predictive equations, Aust. J. Bot., 53, 607–619, 2005. nlme: Linear and Nonlinear Mixed Effects Models, available at: Woinarski, J. C. Z., Risler, J., and Kean, L.: Response of vegetation https://CRAN.R-project.org/package=nlme (last access: 7 Jan- and vertebrate fauna to 23 years of fire exclusion in a tropical Eu- uary 2019), r package version 3.1-137, 2018. calyptus open forest, Northern Territory, Australia, Austral Ecol., Richards, A. E., Andersen, A. N., Schatz, J., Eager, R., Dawes, T. Z., 29, 156–176, 2004. Hadden, K., Scheepers, K., and van der Geest, M.: Savanna burn- Woinarski, J. C. Z., Burbidge, A. A., and Harrison, P. L.: Ongoing ing, greenhouse gas emissions and indigenous livelihoods: Intro- unraveling of a continental fauna: Decline and extinction of Aus- ducing the Tiwi Carbon Study, Austral Ecol., 37, 712–723, 2012. tralian mammals since European settlement, P. Natl. Acad. Sci. Russell-Smith, J., Monagle, C., Jacobsohn, M., Beatty, R. L., Bil- USA, 112, 4531–4540, 2015. bao, B., Millán, A., Vessuri, H., and Sánchez-Rose, I.: Can sa- Wood, B. G., Fogarty, P. J., and Day, K. J.: The land systems of vanna burning projects deliver measurable greenhouse emissions the Darwin region, Northern Territory Conservation Commis- reductions and sustainable livelihood opportunities in fire-prone sion, Northern Territory Conservation Commision, Darwin, NT, settings?, Climatic Change, 140, 47–61, 2013. Australia, available at: http://hdl.handle.net/10070/227809 (last Sankaran, M., Hanan, N. P., Scholes, R. J., Ratnam, J., Augus- access: 4 April 2019), 1985. tine, D. J., Cade, B. S., Gignoux, J., Higgins, S. I., Le Roux, X., Ludwig, F., Ardo, J., Banyikwa, F., Bronn, A., Bucini, G., Caylor, K. K., Coughenour, M. B., Diouf, A., Ekaya, W., Feral, C. J., February, E. C., Frost, P. G. H., Hiernaux, P., Hrabar, H., Metzger, K. L., Prins, H. H. T., Ringrose, S., Sea, W., Tews, J., Worden, J., and Zambatis, N.: Determinants of woody cover in African savannas, Nature, 438, 846–849, 2005. www.biogeosciences.net/16/1493/2019/ Biogeosciences, 16, 1493–1503, 2019
You can also read
